INDEX

Explanations

mentions of things being heavily targeted or heavily emphasized

oai_token-act-pair · gpt-3.5-turbo

words related to overwhelming or excessive experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

slave

-0.73

arel

-0.71

lim

-0.66

olen

-0.65

Ct

-0.65

chal

-0.63

anse

-0.63

ource

-0.62

asury

-0.62

emetery

-0.62

POSITIVE LOGITS

ously

0.69

DERR

0.67

onite

0.66

ĪĴ

0.66

 paste

0.65

ruciating

0.65

 bombard

0.65

ï¸

0.64

ishly

0.64

 Blumenthal

0.61

Activations Density 0.126%

mentions of things being heavily targeted or heavily emphasized

words related to overwhelming or excessive experiences

No Comments

No Known Activations