INDEX

Explanations

words related to misrepresentation or distortion

oai_token-act-pair · gpt-3.5-turbo

instances of distortion or skewing of information

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

cript

-0.79

============

-0.77

APTER

-0.72

dain

-0.70

 Prospect

-0.69

======

-0.69

acht

-0.68

ailable

-0.67

================

-0.66

ept

-0.66

POSITIVE LOGITS

 distort

1.03

 distortions

1.02

 distorted

0.99

 distortion

0.96

usional

0.75

 havoc

0.74

 overest

0.73

roid

0.72

adoes

0.70

 inaccur

0.69

Activations Density 0.035%

words related to misrepresentation or distortion

instances of distortion or skewing of information

No Comments

No Known Activations