INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

mentions of specific individuals, particularly female figures

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iferation

-0.80

WC

-0.72

HCR

-0.69

UST

-0.67

udic

-0.66

 Nanto

-0.64

VL

-0.64

HUD

-0.63

essee

-0.63

BD

-0.63

POSITIVE LOGITS

 Betty

0.92

yip

0.88

keye

0.84

rics

0.82

rand

0.77

hesda

0.76

oro

0.75

rants

0.75

 Seym

0.74

plin

0.72

Activations Density 0.013%

proper nouns

mentions of specific individuals, particularly female figures

No Comments

No Known Activations