INDEX

Explanations

names or terms related to the word "trigger"

oai_token-act-pair · gpt-3.5-turbo

instances of the word fragment "trig" and its variations, indicating a focus on terms related to triggers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Archdemon

-0.76

erest

-0.67

 bount

-0.66

 glim

-0.66

 spotlight

-0.65

 troubled

-0.63

 certific

-0.63

 innocence

-0.63

 rusty

-0.62

 renown

-0.61

POSITIVE LOGITS

quist

0.84

actionDate

0.84

ulence

0.83

opers

0.80

opsis

0.80

itte

0.78

isky

0.78

umatic

0.77

ħĭ

0.76

ilogy

0.75

Activations Density 0.062%

names or terms related to the word "trigger"

instances of the word fragment "trig" and its variations, indicating a focus on terms related to triggers

No Comments

No Known Activations