INDEX

Explanations

locations or organizations related to Los Angeles

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

references to Los Angeles or related geographical terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

urer

-0.72

ury

-0.69

urers

-0.68

erer

-0.66

iller

-0.65

ered

-0.63

iltr

-0.62

rique

-0.62

urn

-0.61

ring

-0.60

POSITIVE LOGITS

nia

0.87

nea

0.87

gee

0.80

say

0.77

cheon

0.77

thus

0.77

dos

0.75

emy

0.75

nam

0.74

NE

0.73

Activations Density 0.252%

locations or organizations related to Los Angeles

references to Los Angeles or related geographical terms

No Comments

No Known Activations