INDEX

Explanations

language related to adequacy and insufficiency

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

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_slefr-ajt/2-res_slefr-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Nichols

-0.82

folk

-0.81

amon

-0.81

destruct

-0.80

hammer

-0.78

dot

-0.78

lyn

-0.75

mas

-0.75

Martin

-0.73

spe

-0.73

POSITIVE LOGITS

ufficient

1.04

 guiActiveUn

0.95

 ILCS

0.94

tarian

0.94

tarians

0.90

Enough

0.89

othing

0.88

ombat

0.87

 Enough

0.86

nces

0.85

Activations Density 0.223%

language related to adequacy and insufficiency

No Comments

No Known Activations