INDEX

Explanations

names related to "Gordon."

oai_token-act-pair · gpt-3.5-turbo

references to individuals named Gordon

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Archdemon

-1.00

atchewan

-0.91

ItemTracker

-0.76

ichita

-0.71

mble

-0.70

atoon

-0.69

ptroller

-0.68

upid

-0.67

ylum

-0.67

hester

-0.66

POSITIVE LOGITS

 Ramsay

1.34

 Gordon

1.00

Gordon

0.98

 Hayward

0.87

 Freeman

0.81

 Yards

0.78

uda

0.75

 Wass

0.74

 Kling

0.69

 Bombay

0.69

Activations Density 0.008%

names related to "Gordon."

references to individuals named Gordon

No Comments

No Known Activations