INDEX

Explanations

words related to intense physical or emotional actions, such as crouching, screeching, or lurking

oai_token-act-pair · gpt-3.5-turbo

words related to crouching and screeching or lurking actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Lutheran

-0.63

ORT

-0.62

orb

-0.61

IELD

-0.59

NRS

-0.59

 Equality

-0.59

Jem

-0.55

 forgiven

-0.54

regn

-0.54

 Aust

-0.54

POSITIVE LOGITS

ches

2.02

ched

1.95

ching

1.92

chers

1.73

cher

1.61

chy

1.52

ch

1.51

kered

1.35

chery

1.31

che

1.22

Activations Density 0.133%

words related to intense physical or emotional actions, such as crouching, screeching, or lurking

words related to crouching and screeching or lurking actions

No Comments

No Known Activations