INDEX

Explanations

phrases or terms related to the usage of similar or identical items, materials, or techniques

oai_token-act-pair · gpt-3.5-turbo

words describing similarities or identical characteristics

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

irable

-0.70

ablishment

-0.69

worthiness

-0.67

 needing

-0.66

bernatorial

-0.65

Ã©n

-0.65

ufact

-0.64

ropolis

-0.63

urd

-0.63

enegger

-0.62

POSITIVE LOGITS

 techniques

1.26

 technique

1.21

 methods

1.11

 terminology

1.10

 pseudonym

1.10

 tactics

1.05

 tactic

1.03

 tricks

1.01

 metaphors

1.00

 pronouns

1.00

Activations Density 0.247%

phrases or terms related to the usage of similar or identical items, materials, or techniques

words describing similarities or identical characteristics

No Comments

No Known Activations