INDEX

Explanations

terms related to technical details and classifications

oai_token-act-pair · gpt-3.5-turbo

terms related to resources and their classifications

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aminer

-0.72

 Canaver

-0.66

ouf

-0.61

zik

-0.56

atta

-0.55

undrum

-0.55

apist

-0.55

ixir

-0.54

amera

-0.54

agogue

-0.54

POSITIVE LOGITS

-,

1.28

and

1.07

 versus

1.05

or

1.02

 equivalents

1.00

 categories

0.97

0.95

0.93

vs

0.90

AND

0.89

Activations Density 0.793%

terms related to technical details and classifications

terms related to resources and their classifications

No Comments

No Known Activations