INDEX

Explanations

phrases related to timing and sequence of events

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

 preserves

-0.66

requires

-0.63

gra

-0.63

ouble

-0.63

adj

-0.62

chron

-0.61

oreal

-0.59

ymes

-0.59

venge

-0.59

 aven

-0.59

POSITIVE LOGITS

 drafting

0.85

 hearing

0.83

 brainstorm

0.76

 designing

0.74

 inception

0.73

 announcing

0.71

 unveiling

0.70

CES

0.70

 questioning

0.69

 deciding

0.69

Activations Density 0.145%

phrases related to timing and sequence of events

No Comments

No Known Activations