INDEX

Explanations

mentions of "new"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "new"

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

 Bash

-0.65

 agon

-0.63

sap

-0.60

 equivalent

-0.60

 precip

-0.59

rig

-0.59

dro

-0.59

 recorded

-0.58

 whipped

-0.58

 vest

-0.58

POSITIVE LOGITS

new

3.79

NEW

2.15

New

1.89

new

1.70

NEW

1.56

fresh

1.33

now

1.31

next

1.23

old

1.21

young

1.14

Activations Density 0.007%

mentions of "new"

instances of the word "new"

No Comments

No Known Activations