INDEX

Explanations

sequences of actions or movements involving people or objects

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

specific continuations after nouns

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

 AssemblyCulture

-0.46

 tänka

-0.42

 abstrato

-0.42

Diweddarwch

-0.42

rungsseite

-0.41

 resourceCulture

-0.40

 godine

-0.39

 pétrole

-0.39

Kritik

-0.37

 Potencia

-0.37

POSITIVE LOGITS

 nakalista

0.69

AndEndTag

0.60

__(/*!

0.56

 CreateTagHelper

0.54

"]();

0.51

BufferException

0.47

ActionCreators

0.45

matchCondition

0.44

rawDesc

0.42

Activations Density 0.056%

sequences of actions or movements involving people or objects

specific continuations after nouns

No Comments

No Known Activations