INDEX

Explanations

numerically ordered information

oai_token-act-pair · gpt-3.5-turbo

references to "one" or singular concepts in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 actionGroup

-0.80

hips

-0.79

osponsors

-0.72

ÃįÃį

-0.69

="#

-0.68

ilty

-0.68

respective

-0.66

ãĤ¦

-0.61

jong

-0.61

oof

-0.61

POSITIVE LOGITS

 hundred

1.05

 Hundred

1.04

 thing

1.02

 Piece

0.97

 drawback

0.88

esan

0.88

 wonders

0.87

 notable

0.86

 reason

0.86

 caveat

0.83

Activations Density 0.077%

numerically ordered information

references to "one" or singular concepts in various contexts

No Comments

No Known Activations