INDEX

Explanations

locations or proper nouns related to "Can"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Can" and its variations, indicating a focus on the theme of ability or possibility

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 anonymity

-0.69

 somew

-0.68

theless

-0.67

 independence

-0.66

etheless

-0.65

 insur

-0.65

 trump

-0.64

 flare

-0.61

 reversal

-0.60

 shortcuts

-0.60

POSITIVE LOGITS

ilon

1.08

avan

0.95

eda

0.92

ab

0.90

ams

0.88

aq

0.88

aston

0.85

alla

0.85

ahl

0.84

amon

0.84

Activations Density 0.059%

locations or proper nouns related to "Can"

occurrences of the word "Can" and its variations, indicating a focus on the theme of ability or possibility

No Comments

No Known Activations