INDEX

Explanations

references to positions and actions related to American football, particularly offensive and defensive linemen

oai_token-act-pair · gpt-3.5-turbo

references to offensive or defensive linemen in football

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ibel

-0.77

topic

-0.75

cause

-0.71

lag

-0.70

safe

-0.70

ointment

-0.67

erd

-0.66

Nav

-0.66

pir

-0.66

travel

-0.66

POSITIVE LOGITS

 lineman

1.23

 linemen

1.15

backer

0.98

 Titans

0.83

 sacks

0.73

 tackles

0.72

 Redskins

0.70

 tackle

0.69

 Marino

0.69

Zig

0.68

Activations Density 0.020%

references to positions and actions related to American football, particularly offensive and defensive linemen

references to offensive or defensive linemen in football

No Comments

No Known Activations