INDEX

Explanations

phrases where something is being criticized or dismissed as unworthy or untrue

oai_token-act-pair · gpt-3.5-turbo

instances of dismissal or rejection regarding various statements or agreements

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

rolet

-0.86

OTAL

-0.73

rol

-0.67

width

-0.67

rone

-0.66

rous

-0.65

ppa

-0.63

PsyNetMessage

-0.62

atl

-0.62

Atk

-0.62

POSITIVE LOGITS

 well

0.98

 opposed

0.91

pired

0.90

 soon

0.89

 follows

0.89

pires

0.84

phy

0.80

criptions

0.79

well

0.78

ylum

0.77

Activations Density 0.147%

phrases where something is being criticized or dismissed as unworthy or untrue

instances of dismissal or rejection regarding various statements or agreements

No Comments

No Known Activations