INDEX

Explanations

phrases indicating different options or choices

oai_token-act-pair · gpt-3.5-turbo

conjunctions, particularly the word "and"

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

asper

-0.87

abul

-0.82

æ©

-0.72

uddy

-0.68

hub

-0.65

Leaks

-0.64

ayson

-0.62

dog

-0.62

pees

-0.62

ibi

-0.61

POSITIVE LOGITS

 thence

0.87

 preferably

0.86

 optionally

0.85

rew

0.81

rogens

0.78

 consequently

0.76

then

0.74

 furthermore

0.73

 then

0.73

 secondly

0.72

Activations Density 0.066%

phrases indicating different options or choices

conjunctions, particularly the word "and"

No Comments

No Known Activations