INDEX

Explanations

phrases indicating duration or time-related references

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

nai

-0.81

otto

-0.76

©¶æ¥µ

-0.74

ãģ®ç

-0.74

 isEnabled

-0.74

IRE

-0.74

\\\\\\\\

-0.73

ettle

-0.73

emale

-0.72

pour

-0.71

POSITIVE LOGITS

 mistakes

0.83

 complications

0.81

 hindsight

0.71

 errors

0.70

 luck

0.68

 those

0.67

 flaws

0.66

 sudden

0.66

 leaks

0.66

 unforeseen

0.65

Activations Density 0.011%

phrases indicating duration or time-related references

No Comments

No Known Activations