INDEX

Explanations

mentions of the term "Pitt" with decreasing positive association

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Pitt" in various contexts

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

eering

-0.69

claimer

-0.67

Copy

-0.66

 Sabha

-0.66

ãĤ°

-0.63

arthed

-0.63

GGGG

-0.63

âĸ¬

-0.62

Øª

-0.62

ATIVE

-0.61

POSITIVE LOGITS

sburgh

1.94

sburg

1.34

Pitt

1.20

 Pitt

1.13

sylvania

0.96

0.85

 Penguins

0.84

sten

0.83

ino

0.82

erman

0.82

Activations Density 0.002%

mentions of the term "Pitt" with decreasing positive association

mentions of the term "Pitt" in various contexts

No Comments

No Known Activations