INDEX

Explanations

identifiers or labels, possibly in a legal or formal context

oai_token-act-pair · gpt-3.5-turbo

references to legal processes or discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ulent

-0.67

ãĥŁ

-0.65

izens

-0.63

 merc

-0.63

@@

-0.60

 sublime

-0.60

 incons

-0.59

çīĪ

-0.58

ne

-0.58

+++

-0.58

POSITIVE LOGITS

 mathemat

0.88

plet

0.88

Dialogue

0.85

yss

0.84

laughs

0.77

 helic

0.75

 VIDE

0.72

velt

0.71

laughter

0.69

 contrace

0.69

Activations Density 1.634%

identifiers or labels, possibly in a legal or formal context

references to legal processes or discussions

No Comments

No Known Activations