INDEX

Explanations

the word "Lind" with the highest activation value possible

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Lind."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

OPLE

-0.85

CLASSIFIED

-0.78

Magikarp

-0.73

Ø©

-0.70

Rated

-0.67

Downloadha

-0.66

ASED

-0.66

avez

-0.66

ABLE

-0.65

ably

-0.63

POSITIVE LOGITS

gren

1.16

strom

1.09

enstein

1.02

seys

0.97

ahl

0.97

isf

0.97

berg

0.96

quist

0.95

hart

0.90

lasses

0.89

Activations Density 0.003%

the word "Lind" with the highest activation value possible

occurrences of the name "Lind."

No Comments

No Known Activations