INDEX

Explanations

parts of the document related to code, including references to specific code segments

oai_token-act-pair · gpt-3.5-turbo

references to code and programming elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Flavoring

-0.81

iatus

-0.76

 Ukrain

-0.76

abouts

-0.74

iscopal

-0.72

 Wynne

-0.71

strous

-0.69

Ń·

-0.69

ESA

-0.69

 Ples

-0.66

POSITIVE LOGITS

 snippet

1.40

 snippets

1.36

base

1.35

velop

1.01

 snipp

1.01

 reuse

0.98

 repositories

0.97

 repository

0.95

unit

0.93

breakers

0.92

Activations Density 0.020%

parts of the document related to code, including references to specific code segments

references to code and programming elements

No Comments

No Known Activations