INDEX

Explanations

large numerical values mentioned in a factual context

oai_token-act-pair · gpt-3.5-turbo

numerical data and statistics related to various reports and metrics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulet

-0.81

sonian

-0.79

sole

-0.78

infeld

-0.69

pler

-0.68

ufact

-0.63

 Bened

-0.63

ESE

-0.63

 fundament

-0.63

 prolong

-0.62

POSITIVE LOGITS

ths

0.87

lbs

0.84

 pounds

0.79

 Â±

0.75

 souls

0.75

 gallons

0.74

sq

0.71

IU

0.71

avg

0.70

 crore

0.69

Activations Density 0.068%

large numerical values mentioned in a factual context

numerical data and statistics related to various reports and metrics

No Comments

No Known Activations