INDEX

Explanations

proper nouns and titles

oai_token-act-pair · gpt-3.5-turbo

mentions of names related to media organizations or publications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

omorphic

-0.63

earable

-0.56

 Procedure

-0.55

govtrack

-0.55

obin

-0.55

 Question

-0.54

HAR

-0.54

isks

-0.51

/(

-0.51

 Detect

-0.51

POSITIVE LOGITS

 favourites

0.85

 countless

0.85

few

0.83

Others

0.83

 favorites

0.82

 innumerable

0.81

 others

0.81

Examples

0.80

 examples

0.79

 many

0.74

Activations Density 0.110%

proper nouns and titles

mentions of names related to media organizations or publications

No Comments

No Known Activations