INDEX

Explanations

written conversations with a colon to indicate a speaker

oai_token-act-pair · gpt-3.5-turbo

punctuation marks or colons at the end of sentences or phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 behavi

-0.80

avorite

-0.72

 rule

-0.72

 caster

-0.69

 depreciation

-0.68

poons

-0.68

respective

-0.66

 reckoning

-0.66

 overl

-0.64

inement

-0.64

POSITIVE LOGITS

 Yeah

1.00

Bye

0.88

Who

0.82

Yes

0.81

Oh

0.81

TBD

0.80

 Impossible

0.77

Huh

0.75

 Okay

0.75

Yep

0.74

Activations Density 0.086%

written conversations with a colon to indicate a speaker

punctuation marks or colons at the end of sentences or phrases

No Comments

No Known Activations