INDEX

Explanations

specific nouns and proper names indicating relationships, actions, or attributes related to people and objects

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

Word before an auxiliary verb

np_acts-logits-general · gemini-2.0-flash

names and roles

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

ViewImports

-0.37

Read

-0.34

Бележки

-0.32

lo

-0.31

eqref

-0.29

ViewBag

-0.29

<eos>

-0.29

Collegamenti

-0.28

…

-0.28

Red

-0.28

POSITIVE LOGITS

bootstrapcdn

0.75

<unused43>

0.74

<unused79>

0.74

[@BOS@]

0.74

<unused14>

0.74

<unused41>

0.74

<unused51>

0.74

<unused23>

0.74

<unused8>

0.74

<unused17>

0.74

Activations Density 0.628%

specific nouns and proper names indicating relationships, actions, or attributes related to people and objects

Word before an auxiliary verb

names and roles

No Comments

No Known Activations