INDEX

Explanations

references to legal proceedings and courtroom activities

oai_token-act-pair · gpt-3.5-turbo

references to juries in legal contexts

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

 Franch

-0.73

irst

-0.68

zl

-0.66

 Occupations

-0.65

 Trip

-0.64

dayName

-0.62

erenn

-0.62

 Patient

-0.62

oms

-0.61

ospital

-0.61

POSITIVE LOGITS

glass

0.84

 jury

0.84

proof

0.78

 verdict

0.77

 jurors

0.77

rooms

0.74

room

0.74

Panel

0.71

ĵĺ

0.70

member

0.69

Activations Density 0.014%

references to legal proceedings and courtroom activities

references to juries in legal contexts

No Comments

No Known Activations