INDEX

Explanations

terms associated with levels or extremes of something, particularly heights and outbreaks

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

redo

-0.78

sbm

-0.62

hemat

-0.61

tan

-0.61

 Merge

-0.60

onto

-0.60

uphem

-0.59

erie

-0.58

true

-0.58

 compass

-0.57

POSITIVE LOGITS

ixties

0.84

 tremend

0.77

ieth

0.76

iest

0.76

era

0.74

 infancy

0.74

osaurs

0.68

itious

0.67

 moment

0.67

 stages

0.66

Activations Density 0.026%

terms associated with levels or extremes of something, particularly heights and outbreaks

No Comments

No Known Activations