INDEX

Explanations

references to locations or elements related to a ranch

oai_token-act-pair · gpt-3.5-turbo

references to ranches and related terms

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

lihood

-0.70

 Hawking

-0.62

ABLE

-0.61

leck

-0.59

 Commonwealth

-0.59

 dism

-0.58

ucket

-0.57

hift

-0.57

rates

-0.56

 decom

-0.56

POSITIVE LOGITS

 ranch

0.95

 Ranch

0.88

 Bundy

0.87

ington

0.80

oÄŁ

0.79

mint

0.77

ing

0.76

oir

0.74

itory

0.73

lands

0.73

Activations Density 0.014%

references to locations or elements related to a ranch

references to ranches and related terms

No Comments

No Known Activations