INDEX

Explanations

numerical sequences in a certain format, possibly related to captions or listings

oai_token-act-pair · gpt-3.5-turbo

numerical data associated with media content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

pole

-0.68

cot

-0.68

antic

-0.63

antis

-0.63

oglu

-0.61

gae

-0.61

ledged

-0.60

leasing

-0.59

mere

-0.59

 whistle

-0.59

POSITIVE LOGITS

 Expand

0.88

 Photos

0.80

 Caption

0.76

 msec

0.75

 Â·

0.74

 è£ıè

0.71

0.70

 Advertisement

0.70

 certs

0.70

=-=-=-=-

0.68

Activations Density 0.065%

numerical sequences in a certain format, possibly related to captions or listings

numerical data associated with media content

No Comments

No Known Activations