INDEX

Explanations

reflections on perceptions and the impact of first impressions

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

atts

-0.71

 challeng

-0.69

 gestation

-0.66

 hips

-0.66

annis

-0.65

vez

-0.62

 contingency

-0.61

 clamp

-0.60

efer

-0.59

foreseen

-0.59

POSITIVE LOGITS

 impression

1.00

able

0.94

 impressions

0.93

eful

0.93

eless

0.93

istic

0.80

ably

0.78

ı

0.77

ability

0.76

tale

0.74

Activations Density 0.009%

reflections on perceptions and the impact of first impressions

No Comments

No Known Activations