INDEX

Explanations

phrases and terms that refer to significant factors influencing various contexts

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

factors and their role

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

chränk

-0.47

ConstraintMaker

-0.45

 passagem

-0.45

̈́

-0.44

 Dunlap

-0.44

Nazionale

-0.44

 mijne

-0.44

 houses

-0.43

 benches

-0.42

 smart

-0.42

POSITIVE LOGITS

 factors

1.01

 Factors

0.95

Factors

0.93

factors

0.85

 FACTORS

0.81

 factores

0.77

 facteur

0.77

 facteurs

0.76

 Faktoren

0.76

 factor

0.71

Activations Density 0.042%

phrases and terms that refer to significant factors influencing various contexts

factors and their role

No Comments

No Known Activations