INDEX

Explanations

apologies or expressions of regret

oai_token-act-pair · gpt-3.5-turbo

instances of the word "sorry" in various contexts

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

Ranked

-0.73

iltration

-0.70

irrel

-0.70

tnc

-0.69

kefeller

-0.68

ernand

-0.68

©¶æ

-0.67

icle

-0.67

acles

-0.66

irled

-0.66

POSITIVE LOGITS

 sorry

1.11

 excuse

0.88

GES

0.84

Sorry

0.83

sorry

0.82

sir

0.80

 guys

0.77

 Sorry

0.74

giving

0.74

tm

0.73

Activations Density 0.024%

apologies or expressions of regret

instances of the word "sorry" in various contexts

No Comments

No Known Activations