INDEX

Explanations

verbs related to observing or monitoring something

oai_token-act-pair · gpt-3.5-turbo

inquiries about future outcomes or scenarios involving comparisons and evaluations

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

ndra

-0.68

 pioneered

-0.66

 innocence

-0.64

 invented

-0.64

otin

-0.60

rir

-0.60

udicrous

-0.57

ools

-0.57

 pity

-0.55

saf

-0.54

POSITIVE LOGITS

 versus

1.14

 compared

1.13

 depends

1.07

 relative

1.01

 varies

0.96

 differs

0.94

 depending

0.93

vis

0.93

 differently

0.88

 differ

0.85

Activations Density 0.376%

verbs related to observing or monitoring something

inquiries about future outcomes or scenarios involving comparisons and evaluations

No Comments

No Known Activations