INDEX

Explanations

positions or roles in a sports context, specifically in football

oai_token-act-pair · gpt-3.5-turbo

references to specific positions in sports, particularly in football

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

isks

-0.75

ingen

-0.68

enos

-0.67

 Warn

-0.66

anz

-0.64

 Jinn

-0.63

Maw

-0.60

 Smile

-0.60

Eff

-0.60

hemy

-0.59

POSITIVE LOGITS

 positions

1.05

 position

1.05

Position

0.86

position

0.85

ession

0.83

ngth

0.83

 Finder

0.82

osition

0.79

 vacated

0.79

 standings

0.77

Activations Density 0.021%

positions or roles in a sports context, specifically in football

references to specific positions in sports, particularly in football

No Comments

No Known Activations