INDEX

Explanations

names of people and places, especially associated with events

oai_token-act-pair · gpt-3.5-turbo

references to individuals and their connections to events or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 hiber

-0.69

Gat

-0.65

 donated

-0.62

 Investig

-0.61

 Entered

-0.60

 starved

-0.59

bur

-0.59

 Epstein

-0.59

 Kingston

-0.57

 inward

-0.56

POSITIVE LOGITS

Trivia

0.86

Tracker

0.81

Conclusion

0.77

References

0.76

.","

0.75

FINEST

0.74

âĵĺ

0.74

Reviewer

0.73

VERTISEMENT

0.73

VAL

0.72

Activations Density 0.458%

names of people and places, especially associated with events

references to individuals and their connections to events or entities

No Comments

No Known Activations