INDEX

Explanations

phrases or words related to writing and data operations

oai_token-act-pair · gpt-3.5-turbo

commands or requests for writing actions in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ĭ±

-0.92

agara

-0.79

Ĥª

-0.68

EGA

-0.68

nels

-0.67

 Ezek

-0.66

 Goodman

-0.65

 disadvant

-0.65

rals

-0.65

 Ukrain

-0.64

POSITIVE LOGITS

lishing

0.74

 aloud

0.72

smanship

0.72

acters

0.72

 instrument

0.71

pee

0.70

RAM

0.70

writing

0.69

 letters

0.69

writ

0.68

Activations Density 0.056%

phrases or words related to writing and data operations

commands or requests for writing actions in various contexts

No Comments

No Known Activations