INDEX

Explanations

reports or studies revealing new information

oai_token-act-pair · gpt-3.5-turbo

statistical reports and findings from studies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

-0.80

lihood

-0.64

viz

-0.59

 Horus

-0.59

Yam

-0.57

Newsletter

-0.55

 Stall

-0.55

 Shall

-0.54

 afore

-0.54

 Kard

-0.53

POSITIVE LOGITS

WASHINGTON

0.87

.–

0.82

Updated

0.74

\.

0.70

*.

0.68

rift

0.66

new

0.64

ename

0.61

PROV

0.61

Activations Density 0.182%

reports or studies revealing new information

statistical reports and findings from studies

No Comments

No Known Activations