INDEX

Explanations

keywords related to attacks or increasing attack power in a gaming context

oai_token-act-pair · gpt-3.5-turbo

terms related to combat mechanics and damage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ©

-0.76

YC

-0.68

zl

-0.66

theless

-0.66

 Vide

-0.63

 Bland

-0.62

 Canucks

-0.61

arton

-0.61

lid

-0.61

milo

-0.60

POSITIVE LOGITS

attack

0.80

 vector

0.74

[+]

0.73

 attack

0.73

ivation

0.72

 against

0.69

iveness

0.69

#$

0.67

ãĥ¼

0.67

ãĤ¹ãĥĪ

0.67

Activations Density 0.034%

keywords related to attacks or increasing attack power in a gaming context

terms related to combat mechanics and damage

No Comments

No Known Activations