INDEX

Explanations

unique number combinations that are potentially related to a specific context or category

oai_token-act-pair · gpt-3.5-turbo

specific numerical values or ratings, particularly in reference to movies or performances

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Whats

-0.72

aido

-0.62

TOP

-0.62

{*

-0.60

dayName

-0.57

 Americ

-0.57

 tsunami

-0.56

sticks

-0.56

SPL

-0.56

iP

-0.56

POSITIVE LOGITS

1.57

1.53

1.52

1.51

1.50

1.49

Activations Density 0.069%

unique number combinations that are potentially related to a specific context or category

specific numerical values or ratings, particularly in reference to movies or performances

No Comments

No Known Activations