INDEX

Explanations

numbered or bulleted lists and significant metrics or indicators

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

words or phrases of conclusion or sequence

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

 snippetHide

-1.00

 queſta

-0.98

WireFormatLite

-0.95

iſchen

-0.93

 الرياضيه

-0.93

 ddelwed

-0.90

ItemBackground

-0.90

extAlignment

-0.88

ſicht

-0.87

<pad>

-0.86

POSITIVE LOGITS

0.39

Finally

0.37

 Finally

0.35

Lastly

0.35

 lastly

0.34

 Lastly

0.33

Activations Density 0.003%

numbered or bulleted lists and significant metrics or indicators

words or phrases of conclusion or sequence

No Comments

No Known Activations