INDEX

Explanations

comparisons of quantities or degrees

oai_token-act-pair · gpt-3.5-turbo

comparative phrases that indicate proportions or likelihoods related to statistical data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Marginal

-0.73

TextColor

-0.68

andise

-0.66

PLA

-0.66

 Beir

-0.65

Pai

-0.65

aria

-0.64

REF

-0.64

 Vegeta

-0.64

SPA

-0.63

POSITIVE LOGITS

 consecut

0.84

agos

0.83

secut

0.81

 sidx

0.81

imeters

0.78

ciating

0.74

xual

0.71

 manif

0.70

uder

0.69

ietal

0.68

Activations Density 0.110%

comparisons of quantities or degrees

comparative phrases that indicate proportions or likelihoods related to statistical data

No Comments

No Known Activations