INDEX

Explanations

terms related to hockey

oai_token-act-pair · gpt-3.5-turbo

references to hockey

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

ibility

-0.96

ible

-0.78

ages

-0.76

ays

-0.71

orative

-0.71

ardless

-0.69

utter

-0.67

AGES

-0.67

iary

-0.66

iger

-0.65

POSITIVE LOGITS

 rink

1.09

 Hockey

1.06

 hockey

0.91

 puck

0.88

ockey

0.85

 Club

0.78

£ı

0.72

 goalie

0.70

 jerseys

0.70

 Cricket

0.67

Activations Density 0.007%

terms related to hockey

references to hockey

No Comments

No Known Activations