INDEX

Explanations

names or titles of judges and legal professionals

oai_token-act-pair · gpt-3.5-turbo

names of judges and legal officials

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

ividual

-0.70

 womb

-0.70

 equivalents

-0.65

 vortex

-0.65

Tokens

-0.65

Spoiler

-0.64

Sensor

-0.62

 tablets

-0.62

 airports

-0.62

 grid

-0.62

POSITIVE LOGITS

ube

0.80

hib

0.78

je

0.76

 Hels

0.74

 Lamb

0.73

ctor

0.73

itus

0.73

 Breed

0.72

iana

0.72

oka

0.72

Activations Density 0.221%

names or titles of judges and legal professionals

names of judges and legal officials

No Comments

No Known Activations