INDEX

Explanations

numerical values and measurements

oai_token-act-pair · gpt-3.5-turbo

articles and nouns related to communication and interaction

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Nept

-0.98

Yan

-0.89

Nan

-0.88

 Brun

-0.84

 Neptune

-0.83

 Thomson

-0.83

 Frey

-0.82

 Synd

-0.81

 Nost

-0.80

 Wong

-0.80

POSITIVE LOGITS

lez

0.97

Loop

0.87

lee

0.87

loop

0.86

endale

0.85

Ģ

0.85

 Loop

0.83

ax

0.82

ace

0.82

LES

0.82

Activations Density 0.320%

numerical values and measurements

articles and nouns related to communication and interaction

No Comments

No Known Activations