INDEX

Explanations

characters or symbols used in internet text slang

oai_token-act-pair · gpt-3.5-turbo

non-standard characters or symbols, indicating possible formatting or encoding issues

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

 Robbins

-0.62

 classified

-0.60

 Hamm

-0.60

 suicide

-0.59

 Dragonbound

-0.59

enegger

-0.59

 Eston

-0.57

 Ferr

-0.57

 simultane

-0.57

 Perkins

-0.57

POSITIVE LOGITS

Ĩ

1.81

ł

1.78

ª

1.77

ĥ

1.76

Į

1.75

©

1.75

¿

1.74

«

1.74

¹

1.74

ŀ

1.73

Activations Density 0.014%

characters or symbols used in internet text slang

non-standard characters or symbols, indicating possible formatting or encoding issues

No Comments

No Known Activations