INDEX

Explanations

prepositions followed by locations or actions

oai_token-act-pair · gpt-3.5-turbo

references to activities and plans involving travel or events

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

most

-0.73

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.72

almost

-0.71

 Approximately

-0.68

ALL

-0.66

illion

-0.64

 FULL

-0.63

 EVERY

-0.63

hest

-0.63

 ONLY

-0.61

POSITIVE LOGITS

 nearby

1.01

Others

0.88

 other

0.86

 later

0.86

 neighbouring

0.84

 others

0.82

 neighboring

0.82

 defunct

0.82

 distant

0.81

 unrelated

0.79

Activations Density 0.754%

prepositions followed by locations or actions

references to activities and plans involving travel or events

No Comments

No Known Activations