INDEX

Explanations

mentions of military organizations and law enforcement agencies

oai_token-act-pair · gpt-3.5-turbo

references to military and law enforcement organizations

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

rencies

-0.82

inges

-0.81

itarian

-0.79

esian

-0.78

ebted

-0.77

ularity

-0.77

cium

-0.77

phal

-0.76

makers

-0.75

izations

-0.75

POSITIVE LOGITS

CCC

0.94

PA

0.93

DEF

0.88

REP

0.85

ENTION

0.85

HAEL

0.84

KK

0.83

CHR

0.81

IL

0.80

LECT

0.80

Activations Density 0.037%

mentions of military organizations and law enforcement agencies

references to military and law enforcement organizations

No Comments

No Known Activations