INDEX

Explanations

proper nouns, specifically names like "Shawn"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Shawn."

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

keley

-0.92

scill

-0.82

gered

-0.75

vernment

-0.72

vernight

-0.72

ught

-0.69

å¥³

-0.67

Ã±a

-0.66

ä»£

-0.65

nces

-0.65

POSITIVE LOGITS

 Michaels

1.12

 Kemp

0.94

sburg

0.90

ees

0.90

ee

0.83

ive

0.81

 Shawn

0.79

 Summers

0.75

 Dillon

0.74

 Thornton

0.74

Activations Density 0.024%

proper nouns, specifically names like "Shawn"

occurrences of the name "Shawn."

No Comments

No Known Activations