INDEX

Explanations

references to behavior, actions, and misconduct related to individuals or groups

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to inappropriate or unacceptable actions and behaviors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

vu

-0.74

opa

-0.67

culus

-0.67

ammy

-0.65

toe

-0.65

CPI

-0.65

nesium

-0.64

ACTED

-0.63

skinned

-0.61

weight

-0.61

POSITIVE LOGITS

 toward

1.16

 exhibited

1.11

 towards

1.08

 patterns

1.07

 modification

1.04

 pattern

0.94

 behavi

0.94

 behaviors

0.94

 habits

0.92

 indicative

0.90

Activations Density 0.107%

references to behavior, actions, and misconduct related to individuals or groups

references to inappropriate or unacceptable actions and behaviors

No Comments

No Known Activations