INDEX

Explanations

references to professors

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the title "Professor" followed by names

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 destro

-0.88

 queen

-0.84

ãĥ¼ãĥ³

-0.79

 cruc

-0.69

 fracture

-0.69

 leash

-0.68

 queens

-0.68

 chorus

-0.67

 takeoff

-0.66

burning

-0.65

POSITIVE LOGITS

essors

1.06

 Laure

0.89

 Puzz

0.89

ĨĴ

0.88

 Emer

0.83

 emer

0.82

 Michel

0.79

 Professor

0.79

umin

0.77

 Wolfgang

0.77

Activations Density 0.021%

references to professors

repeated mentions of the title "Professor" followed by names

No Comments

No Known Activations