INDEX

Explanations

mentions of newsletters

oai_token-act-pair · gpt-3.5-turbo

mentions of newsletters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 reformed

-0.58

 fortun

-0.58

abase

-0.56

 Brist

-0.55

 joints

-0.55

Gaw

-0.54

 Shank

-0.51

 Myst

-0.51

Peb

-0.50

 psychiat

-0.50

POSITIVE LOGITS

interstitial

0.63

ittal

0.62

aceous

0.62

 Continue

0.59

ome

0.58

etsk

0.57

ashington

0.57

ieu

0.57

escent

0.56

ansom

0.56

Activations Density 0.007%

mentions of newsletters

mentions of newsletters

No Comments

No Known Activations