INDEX

Explanations

quotes or direct speech

oai_token-act-pair · gpt-3.5-turbo

references to films and their creators

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

omet

-0.86

utterstock

-0.86

OPER

-0.85

immer

-0.84

 glim

-0.80

 Comcast

-0.76

iver

-0.75

cery

-0.75

atown

-0.74

ormons

-0.73

POSITIVE LOGITS

Sa

1.59

Ba

1.47

Sa

1.43

Za

1.35

Ba

1.29

Ta

1.27

Da

1.23

ta

1.20

Ja

1.19

Cha

1.19

Activations Density 0.381%

quotes or direct speech

references to films and their creators

No Comments

No Known Activations