INDEX

Explanations

phrases related to conditions or consequences of certain actions or beliefs

oai_token-act-pair · gpt-3.5-turbo

instances of the word "don't" or its variations

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

Site

-0.71

 Alleg

-0.68

 Balanced

-0.67

OSP

-0.67

 Policies

-0.64

 Starts

-0.61

Gall

-0.61

 Strategy

-0.60

 Palest

-0.60

inia

-0.59

POSITIVE LOGITS

 bother

0.94

 succeed

0.94

theless

0.85

urtles

0.84

 comply

0.82

 appreciate

0.81

 recognize

0.81

 realize

0.80

 necessarily

0.79

 realise

0.79

Activations Density 0.083%

phrases related to conditions or consequences of certain actions or beliefs

instances of the word "don't" or its variations

No Comments

No Known Activations