INDEX

Explanations

references to the name "Lane" in various contexts

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

iggurat

-0.77

ulatory

-0.74

glomer

-0.74

orate

-0.72

ulates

-0.71

 Seym

-0.70

ģĸ

-0.69

actionDate

-0.69

racted

-0.69

ional

-0.68

POSITIVE LOGITS

 Lane

0.85

backer

0.82

 Bryant

0.75

©¶æ

0.73

ige

0.71

lane

0.70

sey

0.69

forth

0.69

holder

0.67

ham

0.67

Activations Density 0.007%

references to the name "Lane" in various contexts

No Comments

No Known Activations