INDEX

Explanations

mentions of physical and metaphorical heads or leadership positions

oai_token-act-pair · gpt-3.5-turbo

references to "head" or "heads" 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

PsyNetMessage

-1.01

terday

-0.79

DonaldTrump

-0.73

actory

-0.71

arial

-0.67

ancies

-0.66

 mell

-0.66

 conclud

-0.64

pac

-0.64

 resil

-0.63

POSITIVE LOGITS

quarters

1.19

butt

1.12

quartered

1.11

phones

0.98

quarter

0.97

canon

0.97

dress

0.90

lining

0.88

hunter

0.88

line

0.87

Activations Density 0.014%

mentions of physical and metaphorical heads or leadership positions

references to "head" or "heads" in various contexts

No Comments

No Known Activations