INDEX

Explanations

scientific research findings or results

oai_token-act-pair · gpt-3.5-turbo

instances of the word "shown" to indicate evidence or demonstration of a fact

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

etheless

-0.88

ascus

-0.74

ately

-0.69

alez

-0.67

osate

-0.67

idian

-0.67

ilit

-0.65

agan

-0.64

acking

-0.64

claimer

-0.64

POSITIVE LOGITS

ered

1.02

 Thumbnails

0.84

 flashes

0.80

runners

0.78

Reviewer

0.72

¶æ

0.72

 prominently

0.71

 shown

0.70

\\\\\\\\

0.70

shows

0.69

Activations Density 0.043%

scientific research findings or results

instances of the word "shown" to indicate evidence or demonstration of a fact

No Comments

No Known Activations