INDEX

Explanations

occurrences of the word "real" or related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to authenticity and legitimacy

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

 Weston

-0.72

 Reef

-0.66

 Marina

-0.64

 Instruct

-0.62

 Athletics

-0.61

 Fell

-0.60

 Springer

-0.60

 Manny

-0.60

Kro

-0.59

 Chao

-0.58

POSITIVE LOGITS

Â»

1.49

[/

1.34

ãĢį

1.29

\)

1.25

âĢ

1.21

]'

1.20

''

1.18

Â¨

1.18

ãĢı

1.16

Activations Density 0.470%

occurrences of the word "real" or related terms

terms related to authenticity and legitimacy

No Comments

No Known Activations