INDEX

Explanations

phrases indicating conditionality or matters of importance

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

preceding "all," "more," or "course"

np_acts-logits-general · gemini-2.0-flash

after allonce againof course

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 Diſ

-0.98

 Majefty

-0.98

 propOrder

-0.96

 raiſ

-0.95

Vidite

-0.95

SharedDtor

-0.93

ftagPool

-0.91

 themſelves

-0.91

 pleaſure

-0.90

LookAnd

-0.89

POSITIVE LOGITS

pinMode

0.55

0.45

olybden

0.44

el

0.44

<eos>

0.43

éndez

0.43

]").

0.43

0.42

per

0.41

Activations Density 0.298%

phrases indicating conditionality or matters of importance

preceding "all," "more," or "course"

after allonce againof course

No Comments

No Known Activations