INDEX

Explanations

terms related to legal charges and locations they occur in

oai_token-act-pair · gpt-3.5-turbo

structured information regarding charges or legal cases

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

•Partition Features Layer 8 GPT2 Small

" token" vs "Token"

No Comments

Negative Logits

 smugglers

-0.77

 agre

-0.76

 downstream

-0.75

 cris

-0.74

 bloc

-0.74

 retard

-0.74

 trump

-0.74

pse

-0.74

 additionally

-0.73

 sculpt

-0.73

POSITIVE LOGITS

Inv

1.31

Same

1.30

Premium

1.29

Used

1.28

Topic

1.27

Member

1.26

Nothing

1.25

Return

1.25

Topics

1.25

None

1.25

Activations Density 0.184%

terms related to legal charges and locations they occur in

structured information regarding charges or legal cases

No Comments

No Known Activations