INDEX

Explanations

references to specific subjects or entities

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 nhàng

-0.44

 Camden

-0.43

 dere

-0.42

すぎる

-0.41

 lada

-0.41

cej

-0.41

 drag

-0.41

なあ

-0.40

big

-0.40

 useState

-0.40

POSITIVE LOGITS

webElementXpaths

0.66

principalColumn

0.65

IntoConstraints

0.59

 كومونز

0.59

ConstraintMaker

0.57

 disambiguazione

0.56

 Савезне

0.56

 CreateTagHelper

0.56

 nakalista

0.54

 препратки

0.54

Activations Density 0.000%

references to specific subjects or entities

No Comments

No Known Activations