INDEX

Explanations

phrases related to coding or encoding

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "cod."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Prosper

-0.73

 Parish

-0.64

 boycott

-0.63

 independents

-0.62

 boycot

-0.60

 Armenia

-0.59

 charged

-0.58

lain

-0.58

 Negro

-0.57

 Roma

-0.57

POSITIVE LOGITS

ename

1.66

ependent

1.25

epend

1.16

irect

1.15

ependence

1.14

epad

1.11

estones

1.08

ex

1.07

ãĥĥãĥī

1.06

ework

1.00

Activations Density 0.021%

phrases related to coding or encoding

occurrences of the term "cod."

No Comments

No Known Activations