INDEX

Explanations

lexical items related to definitions and descriptions of various terms or concepts

oai_token-act-pair · gpt-3.5-turbo

discussions about definitions and their interpretations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iddles

-0.71

reau

-0.71

ennes

-0.69

liners

-0.66

 Moves

-0.66

ADS

-0.65

DAQ

-0.65

lication

-0.65

llor

-0.65

loads

-0.64

POSITIVE LOGITS

 acceptable

1.01

 permissible

1.01

0.98

 insanity

0.95

 legality

0.91

0.90

"'

0.89

 professionalism

0.88

\"

0.87

 what

0.86

Activations Density 0.228%

lexical items related to definitions and descriptions of various terms or concepts

discussions about definitions and their interpretations

No Comments

No Known Activations