INDEX

Explanations

proper nouns, especially names and affiliations

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

tight

-0.43

slo

-0.42

ult

-0.41

 tight

-0.41

 tightly

-0.40

 Tight

-0.39

Smooth

-0.39

 rút

-0.39

Highly

-0.38

 Momentum

-0.38

POSITIVE LOGITS



0.55

transQ

0.50

erebbe

0.49

 ब्रेकडाउन

0.49

المشاركات

0.49

 protoimpl

0.49

 nonUne

0.49

AutoField

0.48

weightedMode

0.48

 Normdatei

0.47

Activations Density 0.019%

proper nouns, especially names and affiliations

No Comments

No Known Activations