INDEX

Explanations

words related to direct quotes by someone

oai_token-act-pair · gpt-3.5-turbo

references to characters, specifically male and female pronouns

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

Hub

-0.65

 Front

-0.65

 Berk

-0.64

 internationally

-0.64

HG

-0.63

 Royale

-0.62

 Argon

-0.62

 nationally

-0.60

 Batt

-0.60

 Azerb

-0.59

POSITIVE LOGITS

uristic

0.98

zbollah

0.90

eded

0.88

ngth

0.88

 didnt

0.88

mos

0.85

lder

0.85

aped

0.85

gemony

0.85

areth

0.84

Activations Density 0.057%

words related to direct quotes by someone

references to characters, specifically male and female pronouns

No Comments

No Known Activations