INDEX

Explanations

target audiences or levels of expertise related to specific products or content

oai_token-act-pair · gpt-3.5-turbo

terms related to users of varying levels of experience with technology or products

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

 confessions

-0.61

 judgments

-0.61

 slurs

-0.60

gregation

-0.59

 arrests

-0.59

Reloaded

-0.58

 convictions

-0.58

 shroud

-0.56

 Massacre

-0.55

tein

-0.55

POSITIVE LOGITS

 alike

1.44

 wanting

1.23

 wishing

1.23

who

1.13

 interested

1.11

 needing

1.07

Interested

1.00

who

1.00

 aspiring

0.96

 seeking

0.92

Activations Density 0.246%

target audiences or levels of expertise related to specific products or content

terms related to users of varying levels of experience with technology or products

No Comments

No Known Activations