INDEX

Explanations

phrases related to political ideologies and groups

oai_token-act-pair · gpt-3.5-turbo

references to neo-Nazi groups and related ideologies

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

 intervals

-0.73

 exceptions

-0.69

eele

-0.65

AVG

-0.65

*/(

-0.65

 Petty

-0.65

cliffe

-0.65

ulhu

-0.64

ado

-0.64

WD

-0.64

POSITIVE LOGITS

centric

0.97

Georg

0.96

destruct

0.94

Nazi

0.92

driven

0.92

optim

0.91

induced

0.88

analysis

0.88

inspired

0.87

friendly

0.87

Activations Density 0.130%

phrases related to political ideologies and groups

references to neo-Nazi groups and related ideologies

No Comments

No Known Activations