INDEX

Explanations

phrases related to legal actions and admissions of guilt

oai_token-act-pair · gpt-3.5-turbo

instances of the term "guilty" related to legal pleas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

adish

-0.80

edia

-0.76

dos

-0.73

inth

-0.72

 Argent

-0.68

nic

-0.68

inas

-0.66

chn

-0.65

clips

-0.64

care

-0.64

POSITIVE LOGITS

 plea

0.89

 guilty

0.80

icts

0.79

 Guilty

0.79

 verdict

0.69

 repeatedly

0.68

 manslaughter

0.68

ysis

0.68

 voluntarily

0.68

animous

0.67

Activations Density 0.032%

phrases related to legal actions and admissions of guilt

instances of the term "guilty" related to legal pleas

No Comments

No Known Activations