INDEX

Explanations

categories, numeric values, and subjects related to detailed technical or scientific information

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

scientific classifications or names

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

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

 الحره

-0.93

enablog

-0.81

 パンチラ

-0.80

<unused41>

-0.79

<unused28>

-0.79

<unused8>

-0.79

<unused14>

-0.79

[@BOS@]

-0.79

<unused43>

-0.79

<pad>

-0.79

POSITIVE LOGITS

↵↵

0.41

0.33

↵

0.33

0.32

 consists

0.31

DU

0.31

 ис

0.31

Base

0.31

 Base

0.30

This

0.30

Activations Density 0.034%

categories, numeric values, and subjects related to detailed technical or scientific information

scientific classifications or names

No Comments

No Known Activations