INDEX

Explanations

phrases related to ability or incapability

oai_token-act-pair · gpt-3.5-turbo

negations and assertions of ability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Lens

-0.64

 Rutherford

-0.62

 enthusi

-0.62

 Brent

-0.60

IDES

-0.57

 millenn

-0.56

ELD

-0.56

 prompts

-0.55

 conspicuous

-0.55

 Mant

-0.55

POSITIVE LOGITS

't

2.33

NOT

1.32

adian

1.17

 afford

1.14

na

1.06

 hardly

1.05

Â´

1.00

berra

0.99

ny

0.99

 handle

0.96

Activations Density 0.143%

phrases related to ability or incapability

negations and assertions of ability

No Comments

No Known Activations