INDEX

Explanations

action-oriented verbs and infinitives related to functionality or operations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 GreatSchools

-0.58

rboles

-0.55

 đám

-0.54

Климат

-0.52

 gobier

-0.50

bindet

-0.49

rvore

-0.49

 söyledi

-0.48

sistence

-0.47

bersicht

-0.47

POSITIVE LOGITS

 ולה

0.95

\{\\

0.79

Открыть

0.75

lizar

0.72

 oneself

0.71

quisar

0.69

[]

0.69

Estar

0.68

FromArgb

0.68

decer

0.67

Activations Density 0.079%

action-oriented verbs and infinitives related to functionality or operations

No Comments

No Known Activations