INDEX

Explanations

references to statistical analysis and data interpretation

oai_token-act-pair · gpt-3.5-turbo

metrics and rankings related to player performance

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

stal

-0.82

atom

-0.80

fur

-0.73

lene

-0.69

orah

-0.69

mount

-0.66

mun

-0.66

Mos

-0.65

 Brussels

-0.64

 urgently

-0.64

POSITIVE LOGITS

 stats

1.45

 Adjusted

1.43

stats

1.37

Stats

1.37

 metrics

1.32

 STATS

1.29

 Stats

1.24

 batted

1.24

 averages

1.21

 percentile

1.20

Activations Density 0.423%

references to statistical analysis and data interpretation

metrics and rankings related to player performance

No Comments

No Known Activations