INDEX

Explanations

terms related to legal or criminal executions

oai_token-act-pair · gpt-3.5-turbo

references to executions and capital punishment

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

æ©Ł

-0.76

 explan

-0.69

CE

-0.69

thora

-0.67

Fi

-0.66

ECO

-0.66

UTH

-0.66

BuyableInstoreAndOnline

-0.65

 Emerson

-0.65

CRE

-0.65

POSITIVE LOGITS

 inmate

0.99

 inmates

0.87

ifix

0.84

icts

0.83

 prisoners

0.83

 convict

0.81

 executions

0.81

 murder

0.81

quished

0.81

 abolition

0.80

Activations Density 0.078%

terms related to legal or criminal executions

references to executions and capital punishment

No Comments

No Known Activations