INDEX

Explanations

actions of characters in the narrative

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

haven

-0.73

unity

-0.67

ADVERTISEMENT

-0.65

cair

-0.64

Sov

-0.63

 specialization

-0.62

 psychiat

-0.62

KEN

-0.61

Prosecutors

-0.60

Develop

-0.60

POSITIVE LOGITS

 glances

1.03

 gently

0.89

 motions

0.89

 nervously

0.88

 clasp

0.88

 clenched

0.86

 quir

0.84

agall

0.83

rily

0.83

 gest

0.82

Activations Density 0.112%

actions of characters in the narrative

No Comments

No Known Activations