INDEX

Explanations

questions or uncertainties expressed in the text

oai_token-act-pair · gpt-3.5-turbo

questions related to uncertainty and understanding

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

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

ynski

-0.74

WAYS

-0.70

oola

-0.68

ixel

-0.67

cks

-0.66

wal

-0.65

enaries

-0.65

ulkan

-0.65

ortment

-0.65

ainers

-0.64

POSITIVE LOGITS

 else

1.13

 anybody

0.90

 anymore

0.88

 anyone

0.81

 coincidence

0.75

 bothered

0.71

to

0.70

chu

0.67

 exactly

0.66

 anything

0.66

Activations Density 0.095%

questions or uncertainties expressed in the text

questions related to uncertainty and understanding

No Comments

No Known Activations