INDEX

Explanations

indefinite articles 'a' or 'an' followed by an adjective and a noun

oai_token-act-pair · gpt-3.5-turbo

articles and quantifiers describing entities or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Plans

-0.60

 intentions

-0.60

 deals

-0.58

 dealings

-0.57

 agreements

-0.57

 papers

-0.57

 boots

-0.54

 tunnels

-0.54

 unresolved

-0.54

Oracle

-0.54

POSITIVE LOGITS

uras

1.04

unts

1.01

gh

1.00

 subset

1.00

 bunch

0.98

 rarity

0.91

 dime

0.85

 staple

0.85

 couple

0.84

versive

0.84

Activations Density 0.176%

indefinite articles 'a' or 'an' followed by an adjective and a noun

articles and quantifiers describing entities or groups

No Comments

No Known Activations