INDEX

Explanations

words related to key issues, challenges, and conflicts

oai_token-act-pair · gpt-3.5-turbo

key terms related to significant topics or issues

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

uly

-0.64

ADRA

-0.63

lish

-0.62

rocal

-0.61

heid

-0.60

qus

-0.60

guyen

-0.59

uthor

-0.59

plete

-0.58

ruciating

-0.58

POSITIVE LOGITS

 revolves

0.84

 lies

0.82

 involves

0.81

 resides

0.80

 here

0.72

lie

0.72

 relates

0.71

 imaginable

0.71

 besides

0.71

 behind

0.70

Activations Density 0.241%

words related to key issues, challenges, and conflicts

key terms related to significant topics or issues

No Comments

No Known Activations