INDEX

Explanations

verbs or phrases related to causing an action or response

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "prompt" and related terms indicating stimuli or triggers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

gdala

-0.69

 Extrem

-0.69

 Crescent

-0.68

 Sect

-0.68

aples

-0.68

utra

-0.67

hetical

-0.67

lim

-0.66

antam

-0.63

 Scythe

-0.62

POSITIVE LOGITS

 prompt

1.57

 prompts

1.05

 Prompt

0.99

 showc

0.94

itude

0.77

 succession

0.75

 prompting

0.75

shire

0.74

 prompted

0.72

 warnings

0.72

Activations Density 0.005%

verbs or phrases related to causing an action or response

occurrences of the word "prompt" and related terms indicating stimuli or triggers

No Comments

No Known Activations