INDEX

Explanations

the word "longer" paired with various contexts

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a change in status or duration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 opposite

-0.75

imov

-0.71

é¾

-0.68

achev

-0.68

NG

-0.65

EStream

-0.65

hex

-0.64

 teasp

-0.63

println

-0.63

osite

-0.62

POSITIVE LOGITS

 tolerated

0.82

 bothered

0.81

 bothering

0.79

 exists

0.76

 needing

0.75

 needed

0.74

 reliant

0.73

 tolerate

0.73

 able

0.73

 viable

0.72

Activations Density 0.069%

the word "longer" paired with various contexts

phrases indicating a change in status or duration

No Comments

No Known Activations