INDEX

Explanations

mentions of the name "Ellison."

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

worthiness

-0.71

fulness

-0.69

nces

-0.67

ework

-0.66

kered

-0.65

esome

-0.64

eric

-0.64

fully

-0.64

assium

-0.63

ctica

-0.60

POSITIVE LOGITS

 Ellison

0.81

icut

0.72

irez

0.72

mann

0.70

berg

0.69

eering

0.66

urations

0.66

 Franken

0.65

 endorsed

0.64

wald

0.64

Activations Density 0.003%

mentions of the name "Ellison."

No Comments

No Known Activations