INDEX

Explanations

instances of the name "Leonard."

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

skirts

-0.84

glers

-0.80

ħĭ

-0.77

PLA

-0.73

worldly

-0.68

manship

-0.67

selves

-0.67

yrim

-0.66

*/(

-0.65

pron

-0.62

POSITIVE LOGITS

angelo

0.97

 Cohen

0.91

 Leonard

0.91

 Lauder

0.88

ardo

0.85

ency

0.83

 Bernstein

0.82

arious

0.81

Nim

0.80

zman

0.80

Activations Density 0.002%

instances of the name "Leonard."

No Comments

No Known Activations