INDEX

Explanations

names of sports teams, particularly in soccer and football contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

ãĥ©

-0.77

è£ħ

-0.73

Bloomberg

-0.73

%%%%

-0.70

romeda

-0.70

@@@@@@@@

-0.66

artisan

-0.66

cients

-0.65

DonaldTrump

-0.65

 acknow

-0.65

POSITIVE LOGITS

FC

1.18

Ltd

0.84

 Albion

0.84

 rugby

0.80

HS

0.80

 Athletic

0.79

 midfielder

0.78

 Militia

0.77

 Club

0.76

 goalkeeper

0.76

Activations Density 0.056%

names of sports teams, particularly in soccer and football contexts

No Comments

No Known Activations