INDEX

Explanations

legal terms and rulings

oai_token-act-pair · gpt-3.5-turbo

judicial rulings and legal decisions

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

ãĤ§

-0.85

ãĥ¡

-0.81

æ©

-0.76

ãĥ¼ãĥĨãĤ£

-0.71

æ°

-0.70

ãĥ¼ãĥĨ

-0.68

nown

-0.66

Ops

-0.65

 weap

-0.65

POSITIVE LOGITS

 sided

0.90

 justices

0.90

 upholding

0.89

 unanimously

0.89

 verdict

0.88

 exoner

0.86

 rulings

0.86

 precedent

0.83

 invalid

0.82

 Judicial

0.81

Activations Density 0.270%

legal terms and rulings

judicial rulings and legal decisions

No Comments

No Known Activations