INDEX

Explanations

timestamps and other details from a TV schedule or programming guide

oai_token-act-pair · gpt-3.5-turbo

specific symbols or formatting elements typically associated with titles and time references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 licens

-0.85

 denomin

-0.83

 endors

-0.80

 manufacturer

-0.78

 assigned

-0.77

 manufacturers

-0.77

 standardized

-0.76

uid

-0.75

 wage

-0.73

 disadvantage

-0.73

POSITIVE LOGITS

 Panic

1.17

 Revenge

1.16

 Prelude

1.12

 Sleeping

1.11

 Dirty

1.11

 Inferno

1.10

 Apocalypse

1.10

 Nost

1.09

 Silent

1.09

 Salvation

1.08

Activations Density 0.146%

timestamps and other details from a TV schedule or programming guide

specific symbols or formatting elements typically associated with titles and time references

No Comments

No Known Activations