INDEX

Explanations

proper nouns with the word "Kn" in them

oai_token-act-pair · gpt-3.5-turbo

references to the name "Kn" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

quo

-0.82

 Uran

-0.75

pour

-0.74

ional

-0.72

quickShipAvailable

-0.71

REDACTED

-0.71

ORGE

-0.70

DEM

-0.70

ahime

-0.69

anamo

-0.68

POSITIVE LOGITS

uckle

1.09

itting

1.04

ivable

1.03

ows

1.02

ocking

1.02

uckles

0.99

icker

0.94

itty

0.91

itted

0.91

kn

0.90

Activations Density 0.011%

proper nouns with the word "Kn" in them

references to the name "Kn" in various contexts

No Comments

No Known Activations