INDEX

Explanations

words related to influences or consequences

oai_token-act-pair · gpt-3.5-turbo

words related to consequences and influences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 bargain

-0.68

 derby

-0.66

com

-0.66

toy

-0.65

upd

-0.65

timer

-0.64

 preferred

-0.63

rev

-0.62

 sport

-0.62

 sponsor

-0.62

POSITIVE LOGITS

ences

4.88

ence

2.73

encing

2.72

ENCE

2.28

enced

2.27

ENC

1.97

encers

1.92

encer

1.86

encies

1.68

ential

1.68

Activations Density 0.007%

words related to influences or consequences

words related to consequences and influences

No Comments

No Known Activations