INDEX

Explanations

words related to results, outcomes, or consequences

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

kov

-0.69

erness

-0.69

agine

-0.66

 actionGroup

-0.65

resent

-0.64

tera

-0.63

afort

-0.63

tz

-0.62

 Passage

-0.61

conservancy

-0.61

POSITIVE LOGITS

 thereof

0.86

iveness

0.83

ainer

0.79

results

0.74

 result

0.73

ively

0.72

ivity

0.70

result

0.68

 results

0.68

iments

0.68

Activations Density 0.703%

words related to results, outcomes, or consequences

No Comments

No Known Activations