INDEX

Explanations

references to processes and assessments 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

 casualty

-0.62

FACE

-0.60

liest

-0.59

 HELP

-0.59

è¦ļéĨĴ

-0.58

 brace

-0.58

 unaccompanied

-0.57

 nostalg

-0.57

 PRESIDENT

-0.56

 FACE

-0.56

POSITIVE LOGITS

neau

0.87

uese

0.80

anski

0.78

isa

0.78

iy

0.78

combe

0.77

ibo

0.76

anche

0.74

emen

0.73

alion

0.72

Activations Density 0.087%

references to processes and assessments in various contexts

No Comments

No Known Activations