INDEX

Explanations

references or mentions of datasets

oai_token-act-pair · gpt-3.5-turbo

mentions of datasets and related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

odge

-0.80

ogy

-0.72

inence

-0.69

endi

-0.67

ohan

-0.67

pelling

-0.66

ban

-0.66

ingers

-0.65

sterdam

-0.65

ife

-0.64

POSITIVE LOGITS

 dataset

1.09

 datasets

0.89

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.77

ãĤº

0.76

0.73

TPS

0.72

 GOODMAN

0.71

Catal

0.70

catentry

0.69

ãĤ¼ãĤ¦ãĤ¹

0.68

Activations Density 0.021%

references or mentions of datasets

mentions of datasets and related terminology

No Comments

No Known Activations