INDEX

Explanations

statements indicating causality or relationships between concepts

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

which followed by consequence

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

UserScript

-0.55

IsPostBack

-0.47

ponses

-0.46

land

-0.44

sò

-0.43

 Perfon

-0.43

 polig

-0.41

entur

-0.41

enderror

-0.40

assertArray

-0.40

POSITIVE LOGITS

الحياه

0.65

 WHICH

0.63

 Which

0.60

 which

0.57

Which

0.56

which

0.54

AccessorTable

0.53

 mikä

0.50

 vilket

0.50

 MainAxisSize

0.49

Activations Density 0.052%

statements indicating causality or relationships between concepts

which followed by consequence

No Comments

No Known Activations