INDEX

Explanations

adjectives or phrases conveying hidden or underlying meanings

oai_token-act-pair · gpt-3.5-turbo

terms related to political maneuvering and conflict

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

lain

-0.93

SHIP

-0.73

 Indigo

-0.71

pedia

-0.69

 Vance

-0.65

 Caribbean

-0.65

CN

-0.63

FAC

-0.63

Klu

-0.61

ERY

-0.61

POSITIVE LOGITS

 secession

0.97

 thinly

0.80

udos

0.80

umb

0.80

 guiActiveUn

0.80

shr

0.79

ãĥĭ

0.76

isode

0.72

icidal

0.72

iatures

0.71

Activations Density 0.013%

adjectives or phrases conveying hidden or underlying meanings

terms related to political maneuvering and conflict

No Comments

No Known Activations