INDEX

Explanations

phrases that include the word 'said'

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate reported statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 ILCS

-0.89

anu

-0.84

phal

-0.76

otin

-0.76

VIDEO

-0.76

thia

-0.75

gart

-0.74

bart

-0.74

theless

-0.73

Interstitial

-0.72

POSITIVE LOGITS

 goodbye

0.92

doms

0.75

 earlier

0.70

 bluntly

0.70

 markup

0.68

 ahead

0.66

 majorities

0.64

 analysts

0.64

itivity

0.64

 unanimously

0.63

Activations Density 0.240%

phrases that include the word 'said'

phrases that indicate reported statements

No Comments

No Known Activations