INDEX

Explanations

sports-related terms such as leagues and rankings

oai_token-act-pair · gpt-3.5-turbo

references to sports leagues and rankings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

poke

-0.77

anmar

-0.70

tions

-0.70

 comrade

-0.70

rex

-0.68

terday

-0.68

haps

-0.66

ãĤ½

-0.66

enance

-0.66

paio

-0.65

POSITIVE LOGITS

 standings

1.10

 Finals

0.84

 playoffs

0.78

 spotlight

0.72

 lockout

0.72

venth

0.69

 Playoffs

0.67

Era

0.67

 Spotlight

0.66

 hierarchy

0.66

Activations Density 0.086%

sports-related terms such as leagues and rankings

references to sports leagues and rankings

No Comments

No Known Activations