INDEX

Explanations

concepts related to development and strategies within 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

UTF

-0.73

enegger

-0.72

yss

-0.70

onel

-0.67

ocaust

-0.67

psc

-0.66

PRESS

-0.66

Untitled

-0.65

oodle

-0.65

oÄŁ

-0.64

POSITIVE LOGITS

 vary

0.91

 pitfalls

0.88

hooting

0.82

 applicable

0.79

 strategies

0.78

 influencing

0.78

 achievable

0.77

 abound

0.76

 learnt

0.76

ibilities

0.75

Activations Density 0.122%

concepts related to development and strategies within various contexts

No Comments

No Known Activations