INDEX

Explanations

words related to evidence in criminal cases

oai_token-act-pair · gpt-3.5-turbo

references to evidence within a legal context

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

•Evidence

No Comments

Negative Logits

ategory

-0.75

Hop

-0.70

cair

-0.70

orks

-0.68

pop

-0.68

awar

-0.67

cow

-0.67

idays

-0.67

lisher

-0.66

 Wonderful

-0.64

POSITIVE LOGITS

 evidence

1.04

evidence

1.02

 Evidence

0.97

 tampering

0.97

Evidence

0.82

 evid

0.81

edly

0.80

abilia

0.76

 warrants

0.76

ative

0.75

Activations Density 0.018%

words related to evidence in criminal cases

references to evidence within a legal context

No Comments

No Known Activations