INDEX

Explanations

phrases related to legal proceedings and court appearances

oai_token-act-pair · gpt-3.5-turbo

mentions of specific legal courts and their rulings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

yip

-0.82

ãĤ¤ãĥĪ

-0.79

usterity

-0.78

ãĤ£

-0.76

 weap

-0.74

VALUE

-0.72

 contrace

-0.67

ombies

-0.66

ãĥĹ

-0.65

 linem

-0.65

POSITIVE LOGITS

 rulings

1.00

 sided

0.93

 jurisdiction

0.89

 acquitted

0.86

 justices

0.85

 jury

0.84

 subpoena

0.83

 Jury

0.83

 inquest

0.81

 ruling

0.80

Activations Density 0.340%

phrases related to legal proceedings and court appearances

mentions of specific legal courts and their rulings

No Comments

No Known Activations