INDEX

Explanations

elements related to programming and problem-solving methods

oai_token-act-pair · gpt-4o-mini Triggered by @bot

always keep

np_acts-logits-general · gemini-2.5-flash-lite

The neuron detects personal/discourse markers—first- and second-person pronouns and words that signal the speaker's or reader's involvement (e.g., I, we, you, my, our, plan, provide).

oai_token-act-pair · gpt-5-mini Triggered by @vetterc0

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Савезне

-0.49

 Италијани

-0.36

 betweenstory

-0.33

 gradova

-0.33

 defaultstate

-0.32

↹

-0.31

 ویکی‌پدی

-0.31

 Biôgrafia

-0.31

HasAnnotation

-0.29

龇

-0.28

POSITIVE LOGITS

<unused41>

0.85

<unused14>

0.85

<unused3>

0.85

[@BOS@]

0.85

<unused8>

0.85

<unused68>

0.85

<unused52>

0.85

<unused79>

0.85

<unused16>

0.84

<unused23>

0.84

Activations Density 0.760%

elements related to programming and problem-solving methods

always keep

The neuron detects personal/discourse markers—first- and second-person pronouns and words that signal the speaker's or reader's involvement (e.g., I, we, you, my, our, plan, provide).

No Comments

No Known Activations

elements related to programming and problem-solving methods

always keep

The neuron detects personal/discourse markers—first- and second-person pronouns and words that signal the speaker's or reader's involvement (e.g., I, we, you, my, our, plan, provide).

No Comments

No Known Activations