INDEX

Explanations

instances of specific acronyms related to various organizations or events

oai_token-act-pair · gpt-3.5-turbo

mentions of specific organizations or associations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hy

-0.93

Reviewer

-0.87

hod

-0.85

hered

-0.84

mber

-0.83

ty

-0.82

ku

-0.81

chens

-0.79

hide

-0.76

thumbnails

-0.74

POSITIVE LOGITS

 CONC

1.01

 disadvant

0.82

 manif

0.81

 unden

0.80

 challeng

0.79

 suspic

0.79

urrent

0.76

 streng

0.75

ĵĺ

0.75

 describ

0.75

Activations Density 0.011%

instances of specific acronyms related to various organizations or events

mentions of specific organizations or associations

No Comments

No Known Activations