INDEX

Explanations

names of places and people

oai_token-act-pair · gpt-3.5-turbo

sentences that include significant or notable events

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

IFrame

Link

•Partition Features Layer 8 GPT2 Small

" token" vs "Token"

No Comments

Negative Logits

 favor

-0.82

lab

-0.81

 enriched

-0.80

 metic

-0.79

 elbow

-0.78

 indexed

-0.78

 noise

-0.77

 gelatin

-0.77

 regression

-0.77

 sport

-0.76

POSITIVE LOGITS

Imm

1.41

Hope

1.39

Fin

1.30

High

1.29

Wh

1.29

Des

1.28

Nor

1.28

OVER

1.28

Credits

1.27

Fl

1.26

Activations Density 0.257%

names of places and people

sentences that include significant or notable events

No Comments

No Known Activations