INDEX

Explanations

people's names, particularly "Cor"

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the name "Cor."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

éĹĺ

-0.92

CRIP

-0.87

lihood

-0.85

terday

-0.82

anwhile

-0.80

 Dangerous

-0.75

UME

-0.74

TPS

-0.70

 Difficulty

-0.69

WAYS

-0.68

POSITIVE LOGITS

rigan

1.08

fman

1.03

relation

0.96

ruption

0.96

rupt

0.96

bett

0.91

ros

0.90

bis

0.88

porate

0.88

poral

0.87

Activations Density 0.009%

people's names, particularly "Cor"

repeated mentions of the name "Cor."

No Comments

No Known Activations