INDEX

Explanations

references to the act of creating or making something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "created."

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

Thom

-0.73

phia

-0.72

itte

-0.70

 Adin

-0.61

 Mandatory

-0.61

 Treatment

-0.60

itta

-0.59

orie

-0.59

 McGr

-0.59

 Buffett

-0.58

POSITIVE LOGITS

 created

0.89

creat

0.89

 creations

0.86

ãĤ´

0.85

smanship

0.82

ividual

0.79

ngth

0.77

gallery

0.76

srf

0.76

ãĤ´ãĥ³

0.75

Activations Density 0.027%

references to the act of creating or making something

instances of the word "created."

No Comments

No Known Activations