INDEX

Explanations

phrases or sentences indicating a critical or significant moment in a narrative

oai_token-act-pair · gpt-3.5-turbo

references to a "point" in discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 notor

-0.88

 tradem

-0.86

ccording

-0.84

eatures

-0.81

ntil

-0.81

 simul

-0.79

uthor

-0.75

 isot

-0.75

 volunte

-0.73

 actionGroup

-0.73

POSITIVE LOGITS

point

1.19

points

1.09

 point

1.00

 points

0.88

lessly

0.83

Points

0.78

 Point

0.77

blank

0.74

lessness

0.73

ifications

0.72

Activations Density 0.038%

phrases or sentences indicating a critical or significant moment in a narrative

references to a "point" in discussions

No Comments

No Known Activations