INDEX

Explanations

references to actions or measures being taken in response to events or situations

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

ãĥ¼ãĥĨãĤ£

-0.70

gling

-0.65

Fey

-0.64

 Logo

-0.64

orf

-0.63

oiler

-0.62

conservancy

-0.62

 Beir

-0.61

enstein

-0.60

 Cabin

-0.59

POSITIVE LOGITS

able

0.94

 against

0.91

ivism

0.90

ives

0.87

iveness

0.86

 committees

0.85

ional

0.84

aries

0.83

fulness

0.83

ulatory

0.80

Activations Density 0.027%

references to actions or measures being taken in response to events or situations

No Comments

No Known Activations