INDEX

Explanations

statements of knowledge and awareness

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

information and access

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

 endast

-0.57

rungsseite

-0.56

 således

-0.54

 CanadaChoose

-0.54

将其

-0.52

 verwijderen

-0.51

ніципа

-0.51

Personendaten

-0.51

TemporalType

-0.51

 protoimpl

-0.50

POSITIVE LOGITS

 guys

0.60

 gonna

0.56

 stuff

0.56

 thingy

0.54

 thing

0.53

somebody

0.50

 somebody

0.50

 outta

0.50

 weird

0.49

 pretty

0.49

Activations Density 0.220%

statements of knowledge and awareness

information and access

No Comments

No Known Activations