INDEX

Explanations

references to the soccer club Manchester United

oai_token-act-pair · gpt-3.5-turbo

references to Manchester United

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Els

-0.90

Narr

-0.81

ãĥ´

-0.80

 Ples

-0.79

 Macy

-0.73

natureconservancy

-0.73

ãĤ¦

-0.73

ajor

-0.72

Zup

-0.71

ãĤ·ãĥ£

-0.71

POSITIVE LOGITS

 footballer

0.96

 Hots

0.95

 Football

0.89

FC

0.89

 striker

0.86

 sacked

0.86

liga

0.84

 Derby

0.82

 goalkeeper

0.79

 midfielder

0.79

Activations Density 0.070%

references to the soccer club Manchester United

references to Manchester United

No Known Activations