INDEX

Explanations

mentions of actions or instructions

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

ndra

-0.85

ambo

-0.82

tions

-0.80

-+-+

-0.77

ntil

-0.70

Ü

-0.68

nell

-0.65

otten

-0.64

aza

-0.64

ategories

-0.64

POSITIVE LOGITS

 stride

1.02

 cues

1.02

 seriously

1.00

 plunge

0.97

 reins

0.96

cue

0.95

 liberties

0.92

 virginity

0.88

 aback

0.86

 lightly

0.84

Activations Density 1.518%

mentions of actions or instructions

No Comments

No Known Activations