INDEX

Explanations

the word "brief"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "brief."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

rogens

-0.72

ceans

-0.71

 Cosponsors

-0.70

ALD

-0.67

uben

-0.64

IRO

-0.62

ilon

-0.62

rees

-0.61

obal

-0.61

OURCE

-0.60

POSITIVE LOGITS

case

1.52

cases

1.40

 stint

0.95

er

0.95

 foray

0.92

 glimps

0.90

 resp

0.88

 interruption

0.87

 glimpse

0.87

 hiatus

0.85

Activations Density 0.039%

the word "brief"

occurrences of the word "brief."

No Comments

No Known Activations