INDEX

Explanations

words related to topography or geographical features, especially ones related to hillsides or cliffs

oai_token-act-pair · gpt-3.5-turbo

references to the theme of embarrassment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hips

-0.81

ergy

-0.79

RAM

-0.76

TON

-0.76

gem

-0.74

Bey

-0.73

KEY

-0.72

STEM

-0.72

ELY

-0.70

Labor

-0.70

POSITIVE LOGITS

emb

1.31

ourgeois

0.94

arding

0.87

anked

0.79

 hanged

0.78

bourg

0.76

 showc

0.75

ounter

0.75

olicy

0.75

umm

0.74

Activations Density 0.006%

words related to topography or geographical features, especially ones related to hillsides or cliffs

references to the theme of embarrassment

No Comments

No Known Activations