INDEX

Explanations

iterations or coding-related terms

oai_token-act-pair · gpt-3.5-turbo

instances of the term "iter."

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

ership

-0.71

¥µ

-0.70

 sudden

-0.68

 blind

-0.66

 bottled

-0.66

 blinded

-0.65

orld

-0.64

 ceremonies

-0.62

ruciating

-0.61

FACE

-0.61

POSITIVE LOGITS

iter

1.39

mination

0.89

aton

0.86

asury

0.86

ason

0.85

itar

0.81

kson

0.75

itance

0.74

idis

0.74

itation

0.74

Activations Density 0.006%

iterations or coding-related terms

instances of the term "iter."

No Comments

No Known Activations