INDEX

Explanations

references to vertical and horizontal orientations in 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

giving

-0.82

mberg

-0.80

REDACTED

-0.80

ãģ®éŃĶ

-0.80

unes

-0.76

ãģ¦

-0.74

IVERS

-0.74

erv

-0.73

nil

-0.72

lyak

-0.72

POSITIVE LOGITS

 axis

1.02

 stripes

0.92

 separation

0.91

 takeoff

0.89

 ascent

0.88

 dimension

0.87

 orientation

0.86

 leap

0.86

 shaft

0.86

 vertical

0.85

Activations Density 0.007%

references to vertical and horizontal orientations in various contexts

No Comments

No Known Activations