INDEX

Explanations

occurrences of errors

oai_token-act-pair · gpt-3.5-turbo

instances of the word "occurred" indicating events or incidents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

axter

-0.88

step

-0.79

iled

-0.79

arta

-0.76

ierre

-0.76

ilts

-0.75

arching

-0.72

este

-0.72

tier

-0.71

obb

-0.71

POSITIVE LOGITS

 occurring

0.89

occ

0.81

rences

0.75

ĸļ

0.75

TextColor

0.74

 undone

0.73

cffffcc

0.73

 occurred

0.73

uates

0.72

 occurrences

0.71

Activations Density 0.014%

occurrences of errors

instances of the word "occurred" indicating events or incidents

No Comments

No Known Activations