INDEX

Explanations

references to time-related durations or intervals

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

surv

-0.64

inav

-0.57

irie

-0.57

 Sack

-0.55

XII

-0.54

missions

-0.52

SPR

-0.51

Maker

-0.51

huge

-0.51

Fort

-0.51

POSITIVE LOGITS

 attribution

0.61

£ı

0.57

 dash

0.56

 glare

0.55

aga

0.54

 preferred

0.54

ashington

0.52

agos

0.52

 elapsed

0.50

poke

0.49

Activations Density 0.034%

references to time-related durations or intervals

No Comments

No Known Activations