INDEX

Explanations

phrases indicating belief or assumption

oai_token-act-pair · gpt-3.5-turbo

the repetition of the word "believed."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

agher

-0.84

 Lonely

-0.71

nec

-0.68

ingers

-0.62

ateral

-0.62

plan

-0.61

Interstitial

-0.60

aste

-0.59

adan

-0.59

effect

-0.59

POSITIVE LOGITS

 believed

0.85

rul

0.80

rill

0.80

ieved

0.79

ieve

0.76

 believes

0.74

 believe

0.73

GGGGGGGG

0.73

 believing

0.70

 Tradable

0.69

Activations Density 0.010%

phrases indicating belief or assumption

the repetition of the word "believed."

No Comments

No Known Activations