INDEX

Explanations

phrases related to actions or events happening inside a house

oai_token-act-pair · gpt-3.5-turbo

references to locations, specifically homes and houses

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

ansen

-0.75

rencies

-0.70

emonic

-0.70

andum

-0.68

xit

-0.67

ename

-0.67

iferation

-0.66

ioxide

-0.65

 gloss

-0.63

undo

-0.63

POSITIVE LOGITS

nikov

0.82

©¶æ

0.82

 courtyard

0.80

bush

0.79

yard

0.77

 burglary

0.77

stead

0.76

 raided

0.75

owner

0.74

chool

0.74

Activations Density 0.094%

phrases related to actions or events happening inside a house

references to locations, specifically homes and houses

No Comments

No Known Activations