INDEX

Explanations

names related to media personalities or characters (Kopitar, Moria vs Djurkovic, etc.)

oai_token-act-pair · gpt-3.5-turbo

names and terms associated with specific people or characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

OUP

-0.77

DragonMagazine

-0.75

ANCE

-0.74

ciating

-0.74

IZE

-0.70

ATHER

-0.70

MG

-0.66

ãĤĭ

-0.66

IGHTS

-0.65

aceous

-0.65

POSITIVE LOGITS

fw

0.85

inia

0.80

tz

0.80

tle

0.77

wana

0.73

asar

0.70

rael

0.70

anyahu

0.70

itsch

0.69

oro

0.69

Activations Density 0.072%

names related to media personalities or characters (Kopitar, Moria vs Djurkovic, etc.)

names and terms associated with specific people or characters

No Comments

No Known Activations