INDEX

Explanations

references to specific geographical locations or notable sites

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

SourceChecksum

-0.45

 murale

-0.38

Datuak

-0.36

 Oren

-0.35

 audience

-0.34

 supp

-0.34

 graphique

-0.33

unger

-0.33

 SOLD

-0.33

SOLD

-0.32

POSITIVE LOGITS

 <=",

0.73

 offshore

0.52

 zone

0.46

ManagedBean

0.45

expandindo

0.44

ScopeManager

0.44

RTGC

0.43

zone

0.43

 DbType

0.42

 steered

0.42

Activations Density 0.058%

references to specific geographical locations or notable sites

No Comments

No Known Activations