INDEX

Explanations

references to attributes in programming contexts, particularly related to data structures and frameworks

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

 Mendes

-0.66

 Pali

-0.61

 Mund

-0.60

 Bergmann

-0.60

 Oliva

-0.60

inden

-0.59

 Gough

-0.58

ide

-0.57

ibe

-0.57

owe

-0.56

POSITIVE LOGITS

attr

1.00

 attr

0.92

 Attr

0.89

Attr

0.84

Attrs

0.69

 auroit

0.63

ATTR

0.63

 attrs

0.63

 desmotivaciones

0.63

 ATTR

0.60

Activations Density 0.015%

references to attributes in programming contexts, particularly related to data structures and frameworks

No Comments

No Known Activations