INDEX

Explanations

descriptions related to sports teams' performance and standings

oai_token-act-pair · gpt-3.5-turbo

references to sports seasons and team performance metrics

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

 Cosponsors

-0.93

 Actor

-0.79

 authorised

-0.77

Actor

-0.75

 Gael

-0.74

 Activ

-0.74

 Whilst

-0.73

 concess

-0.72

 Anon

-0.72

ech

-0.72

POSITIVE LOGITS

 playoff

1.76

 playoffs

1.56

 postseason

1.50

 Playoff

1.49

 Playoffs

1.34

 wins

1.26

 undefeated

1.22

 championships

1.22

 championship

1.21

 matchups

1.21

Activations Density 0.592%

descriptions related to sports teams' performance and standings

references to sports seasons and team performance metrics

No Comments

No Known Activations