INDEX

Explanations

entities associated with specific keywords or acronyms related to various categories or disciplines

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

ható

-0.59

dollis

-0.57

 aikana

-0.56

 noastre

-0.54

PRESA

-0.54

Coordenadas

-0.53

 picioare

-0.53

vábbi

-0.53

беріга

-0.53

ásra

-0.52

POSITIVE LOGITS

tt

0.89

0.88

tet

0.85

tttt

0.84

ttt

0.82

at

0.81

ET

0.78

ty

0.71

zt

0.71

et

0.70

Activations Density 0.440%

entities associated with specific keywords or acronyms related to various categories or disciplines

No Comments

No Known Activations