INDEX

Explanations

words related to arguments or debates

oai_token-act-pair · gpt-3.5-turbo

linguistic forms that denote actions or characteristics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ilities

-0.64

 Seym

-0.63

ulates

-0.62

raints

-0.61

ility

-0.60

ĳå£«

-0.60

ulating

-0.59

ADRA

-0.59

SU

-0.59

 ordinary

-0.58

POSITIVE LOGITS

oad

1.08

ength

1.00

oaded

0.93

gling

0.90

ibrary

0.89

uci

0.88

ogue

0.86

phrine

0.78

erie

0.77

xual

0.77

Activations Density 0.105%

words related to arguments or debates

linguistic forms that denote actions or characteristics

No Comments

No Known Activations