INDEX

Explanations

words related to limits or restrictions

oai_token-act-pair · gpt-3.5-turbo

terms related to limits or restrictions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

avi

-0.79

otive

-0.78

amara

-0.75

agent

-0.71

ee

-0.70

thora

-0.69

indust

-0.68

ives

-0.67

iquette

-0.67

Reply

-0.66

POSITIVE LOGITS

 capped

0.99

 caps

0.73

 pegged

0.72

 compens

0.68

stan

0.68

 Dunn

0.65

 nickel

0.65

llan

0.65

locked

0.63

ULAR

0.62

Activations Density 0.012%

words related to limits or restrictions

terms related to limits or restrictions

No Comments

No Known Activations