INDEX

Explanations

numerical measurements, potentially related to technology or scientific data

oai_token-act-pair · gpt-3.5-turbo

references to numerical values or data points

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

urus

-0.61

 neighb

-0.59

 temptation

-0.58

book

-0.58

 gifted

-0.57

 stagnation

-0.56

 Chronicle

-0.56

entry

-0.56

ide

-0.56

hers

-0.55

POSITIVE LOGITS

1.27

%-

0.97

0.96

0.93

rpm

0.92

0.87

%,

0.86

ACP

0.86

nm

0.85

Activations Density 0.108%

numerical measurements, potentially related to technology or scientific data

references to numerical values or data points

No Comments

No Known Activations