INDEX

Explanations

the word "finished" with varying activation levels

oai_token-act-pair · gpt-3.5-turbo

instances of the word "finished" used in various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

elled

-0.78

arta

-0.78

ogy

-0.77

alle

-0.75

eros

-0.75

ships

-0.75

utra

-0.74

oland

-0.74

ational

-0.74

amount

-0.72

POSITIVE LOGITS

 Finished

0.96

 finished

0.91

 Finish

0.82

 finishing

0.81

 finish

0.80

 unfinished

0.73

Cec

0.71

 finishes

0.70

 Gael

0.69

ÃįÃį

0.68

Activations Density 0.010%

the word "finished" with varying activation levels

instances of the word "finished" used in various contexts

No Comments

No Known Activations