INDEX

Explanations

statements related to investigations or investigators

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of "investigator" or "investigators" in the context of inquiry or investigation

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

yip

-0.79

jri

-0.76

lord

-0.70

church

-0.70

yards

-0.69

birth

-0.67

eworld

-0.65

 Fiesta

-0.64

Pyr

-0.63

pop

-0.63

POSITIVE LOGITS

 probing

1.15

 investigating

1.04

 investigator

0.92

 subpoen

0.92

 uncover

0.88

 examining

0.86

 Grimoire

0.84

 Investigations

0.83

 interrog

0.82

 investigation

0.82

Activations Density 0.032%

statements related to investigations or investigators

repeated mentions of "investigator" or "investigators" in the context of inquiry or investigation

No Comments

No Known Activations