INDEX

Explanations

the concept of rigging or manipulation

oai_token-act-pair · gpt-3.5-turbo

mentions of systems or processes being "rigged."

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

lication

-0.77

exist

-0.73

gap

-0.72

autions

-0.72

clamation

-0.70

severe

-0.69

chal

-0.69

grave

-0.67

eger

-0.67

older

-0.67

POSITIVE LOGITS

 rigged

1.43

 rigging

1.35

eering

1.01

 crooked

0.78

 gambling

0.76

 betting

0.74

 rotten

0.74

 cheating

0.71

rig

0.70

 reel

0.69

Activations Density 0.018%

the concept of rigging or manipulation

mentions of systems or processes being "rigged."

No Comments

No Known Activations