INDEX

Explanations

sentences with question marks indicating a query or uncertainty

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

places

-0.68

chairs

-0.67

ipel

-0.65

runners

-0.65

gra

-0.64

ilan

-0.64

chat

-0.61

forts

-0.60

agra

-0.59

 mosqu

-0.59

POSITIVE LOGITS

 Thou

0.87

 thou

0.82

 anybody

0.78

 nobody

0.73

 anyone

0.72

 there

0.70

)),

0.67

fortunately

0.66

YOU

0.66

it

0.65

Activations Density 0.052%

sentences with question marks indicating a query or uncertainty

No Comments

No Known Activations