INDEX

Explanations

past tense verbs and terms associated with activation states

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

verbs followed by particles

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

AlterField

-0.48

of

-0.42

the

-0.38

LikeLike

-0.38

 despite

-0.36

AddField

-0.35

mobileqq

-0.35

 with

-0.35

 only

-0.34

POSITIVE LOGITS

majánló

0.77

ientras

0.69

 '\\;'

0.66

 dezelve

0.63

 ainfi

0.63

хьтан

0.62

 zoude

0.62

IBLIO

0.61

 queſta

0.61

étoit

0.60

Activations Density 0.054%

past tense verbs and terms associated with activation states

verbs followed by particles

No Comments

No Known Activations